A novel e-commerce customer continuous purchase recommendation model research based on colony clustering

Customer purchasing behaviour in e-commerce platform has become uncertainty and jump affected by the contexts. Existing personalised recommendation models failed to deal with the problem well and they cause loss of customers constantly. This paper puts forward a novel model based on ant colony clustering algorithm to improve customers' continuous purchase intention, including for customer interest is not drift and interest has already shifting. Firstly, for the research field of e-commerce, it gives definition and structured expression to contexts connotation. Secondly, for the problem that growing data sparseness in recommendation system, it introduces ant colony algorithm to cluster similar users in order to reduce the number of candidate neighbour sets and user similarity computing time. Thus, it can improve the target users of nearest neighbour search accuracy. On this basis, according to customer interest variation characteristics, we put forward the dynamic collaborative filtering recommendation alg...